Declan Walsh, MD

We have known for some time that initial encouraging results from some chemotherapy studies have not been borne out when the therapy is introduced into wider use in the real world. This study raises the same issue except in relation to adverse events. Importantly, it examines the issue among people treated with immunotherapy — PD-1 and PD-L1 checkpoint inhibitors — for NSCLC, most of whom had prior standard chemotherapy. It also is an example of big data research in 2,800 patients identified from the OptumLabs Data Warehouse.
A higher prevalence of adverse events was found than in the initial trials of these agents. The authors note the prior experience of arthralgias associated with aromatase inhibitors found years later to be sixfold more common than early reports suggested. So, the issue is relevant to antitumor efficacy and toxicity, and not just for traditional hormonal or chemotherapy but also for immunotherapy. Perhaps it is to be expected with the introduction of any new antitumor therapy into the wider population. Patients on paper may resemble those who participated in early investigations but are actually subtly different with regard to co-morbidities, or nutritional or performance status. Nevertheless, it is important to detect these issues early in any new drug’s life cycle.
Most toxicities associated with immunotherapies reported to date are qualitatively and quantitatively different than those reported with cytotoxic drugs. The agents are generally well tolerated but have a different adverse event profile. Although relatively uncommon, these toxicities are multisystem in nature and include endocrine, eye, heart, kidney and liver problems, many of them being some form of “-itis.” This study showed that hypophysitis — or pituitary inflammation — was four times more common than earlier reported. There also is a potential timing issue; it is possible that such problems might take years to become evident. The increased incidence of lymphomas years after kidney transplantation immunosuppression comes to mind.
The challenge, as we hope to both cure more cancers and significantly extend life expectancy for others, is how to improve monitoring not just for efficacy but also of toxicity. Extension of life without an accompanying increase in the quality of life may not be the best outcome. In the setting of a treatable but incurable disease, this exposes that patient to multiple long-term morbidities, related to both the disease and treatment. In the immunotherapy era, perhaps we will need an expansion of the multidisciplinary team to include new chronic toxicity management clinics staffed by endocrinologists, cardio-oncologists and others. In those who are cured, survivorship plans and services in particular will need to expand, and they must be evidence-based and protocol-driven.
One limitation of this study was that the authors were unable to account for insurance status. This often is the group with more co-morbidities and worse outcomes. It is possible that the adverse events identified might be seen in a different light or have a different profile if these data were known. Immunotherapy advances are dramatic and welcome. However, based on these data, in a small percentage of patients they risk adverse events with acute and chronic multisystem organ dysfunction. These clinical toxicities may turn out to be subtle, chronic, delayed and cumulative. New structured approaches to toxicity monitoring and management seem appropriate to take full advantage of these exciting new agents. It is important to remember that clinicians often underestimate patients’ true suffering. Better symptom profiling and early involvement of supportive and palliative care services has been shown to be effective and must be part of the response in a new era of rapid advances in immunotherapy.